Search

In structural health monitoring (SHM) and control, the structure can be instrumented with an array of sensors forming a redundant sensor network, which can be utilized in sensor fault diagnosis. In this study, the objective is to detect, identify, and quantify a sensor fault using the structural response data measured with the sensor network. Seven different sensor fault types are investigated and modelled: bias, gain, drifting, precision degradation, complete failure, noise, and constant with noise. The sensor network is modelled as a Gaussian process and each sensor in the network is estimated in turn using the minimum mean square error (MMSE) estimation The sensor fault is identified and quantified using the multiple hypothesis test utilizing the generalized likelihood ratio (GLR). The proposed approach is experimentally verified with an array of accelerometers assembled on a wooden bridge. Different sensor faults are simulated by modifying a single sensor. The method is able to detect a sensor fault, identify and correct the faulty sensor, as well as identify and quantify the fault type.

This paper investigates the problem of fault estimation and fault-tolerant control against sensor failures for a class of nonlinear Itô stochastic systems with simultaneous input and output disturbances. By using a new descriptor sliding mode approach, an accurate estimation of the system states, fault vector and disturbances can be obtained simultaneously. Based on the state estimates, an integral-type sliding mode control scheme against faults and disturbances is proposed to stabilize the resulting fault system. It is shown that the reachabilities of the proposed sliding mode surfaces can be guaranteed in both the state estimate space and the estimation error space simultaneously under the designed control schemes. Finally, a numerical example is presented to illustrate the effectiveness and applicability of the proposed technique. Refereed/Peer-reviewed

In this paper, an online sensor fault detection and isolation (FDI) scheme is proposed based on an emerging Artificial Immune System (AIS) algorithm, namely Dendritic Cell Algorithm (DCA). Our proposed methodology is utilized in a distributed manner in order to perform sensor FDI in complex systems. The proposed methodology is then applied to a wind turbine benchmark model in order to demonstrate its capabilities. 2016 IEEE. Qatar Foundation; Qatar National Research Fund Scopus

This paper presents an active Fault Tolerant Control (FTC) strategy for induction motor (IM) that ensures Field Oriented Control (FOC) and offset the effect of the sensor faults despite of the load torque disturbance. The proposed approach uses a fuzzy descriptor approach to estimate simultaneously the system state and the sensor fault. The physical model of IM is approximated by the Takagi-Sugeno (T-S) fuzzy technique in the synchronous d-q rotating frame with field-oriented control strategy. The stability conditions are analyzed using Lyapunov theory. The controller and observers gains are calculated by solving a set of Linear Matrix Inequalities (LMIs). Finally, the effectiveness of the proposed strategy have been illustrated in simulation and experimental results., U ovom radu je predstavljena strategija upravljanja otpornog na kvarove za asinkroni motor koja omogućuje vektorsko upravljanje bez pogreške uslijed kvara senzora i postojećeg poremećaja momenta tereta. Predloženi pristup koristi neizraziti deskriptor za estimaciju stanja sustava i kvara senzora. Fizikalni model asinkronog motora s vektorskim upravljanjem aproksimiran je korištenjem Takagi-Sugeno modela u rotirajućem d-q koordinatnom sustavu. Uvjeti stabilnosti analizirani us korištenjem Ljapunovljeve teorije. Konstante pojačanja regulatora i obzervera su izračunati rješavanjem skupa linearnih matričnih nejednadžbi. Učinkovitost predložene strategije je ilustrirana simulacijskim i eksperimentalnim rezultatima.

This paper presents a effective sensor fault detection and system reconfiguration approach for DC-DC Boost converter. We consider to design a Luenberger observer to solve the problem of the sensor fault detection of the DC-DC Boost converter. We establish mathematical model according to the state of the switch . Luenberger observer is designed to produce residual errors and analyse faults. We detect three types of sensor faults online and compare residuals with thresholds. The system is able to maintain stability by taking system reconfiguration approach. The effectiveness of this approach is demonstrated by simulations.

The hard support vector regression attracts little attention owing to the overfitting phenomenon. Recently, a fast offline method has been proposed to approximately train the hard support vector regression with the generation performance comparable to the soft support vector regression. Based on this achievement, this article advances a fast online approximation called the hard support vector regression (FOAHSVR for short). By adopting the greedy stagewise and iterative strategies, it is capable of online estimating parameters of complicated systems. In order to verify the effectiveness of the FOAHSVR, an FOAHSVR-based analytical redundancy for aeroengines is developed. Experiments on the sensor failure and drift evidence the viability and feasibility of the analytical redundancy for aeroengines together with its base—FOAHSVR. In addition, the FOAHSVR is anticipated to find applications in other scientific-technical fields.

This paper presents a study of current sensor failure in a Direct Torque Control applied to a Double Fed Induction Generator based Variable Speed Wind System. The effect of scaling and offset current sensor errors is discussed through sensibility analysis. A control reconfiguration is then proposed to remedy this sensor failure. Simulation results emphasize the good performances of the proposed current sensor fault tolerant control

Zmanjševanje porabe energije spada med najbolj pereče globalne izzive današnjega časa. Za ogrevanje zgradb se porabi približno 40 % svetovne proizvodnje energije. Med največje porabnike energije v zgradbah spadajo sistemi za ogrevanje, prezračevanje in klimatizacijo (angl. Heating, Ventilation and Air Conditioning, HVAC). Za varčno porabo energije je tako potrebno zagotoviti zanesljivo delovanje HVAC sistema. Pri načrtovanju vodenja HVAC sistemov je treba upoštevati, da se lahko vsak sestavni del opreme pokvari. Izpadi posameznih elementov lahko povečajo porabo energije, skrajšajo življenjsko dobo ostale opreme in praviloma zmanjšajo bivalno udobje. Izpad posameznih merilno-regulacijskih členov v sistemu, pri vodenju temperature, povzroči neželene spremembe v celotni regulacijski zanki, saj regulator ne dobi pravilen odčitek senzorja. Ob izpadu merilnega člena v nekem prostoru je običajna praksa, da ''zamrznemo'' izhod iz ustreznega regulatorja. Le-to pa ni vedno najbolj ustrezna rešitev, ker lahko s tem povzročimo znatna odstopanja temperature v prostoru. Problem, ki smo ga obravnavali v diplomskem delu, je razviti in testirati algoritem, ki bo uspešno nadomestil izpadli merilni člen. Omenjen algoritem mora biti enostaven in primeren za implementacijo v praksi. V diplomskem delu je predlagan inovativni koncept za določanje odprtosti izvršnega člena ob izpadu senzorja. Temelji na ugotovitvi, da so odprtosti ventilov v sosednjih prostorih in dejavniki, ki vplivajo na temperaturo posameznih prostorov, medsebojno korelirani. Zato vpeljemo nove funkcije, ki odražajo medsebojna razmerja odprtosti ventilov sosednjih prostorov. Pri izpadu določenega senzorja, predlagani sistem vklopi regulacijsko strukturo za vodenje temperature na osnovi izračunanih funkcij razmerij odprtosti ventilov. Algoritem smo preiskusili na simulacijskem modelu prostora, zgrajenem v programskem okolju MATLAB/Simulink. Za potrebe simulacije smo uporabili realne vremenske podatke za mesto Ljubljana za obdobje enega leta, s frekvenco vzorčenja ene ure. Rezultati so pokazali veliko ujemanje med sistemom brez izpadov in sistemom z odpovedjo enega merilnega člena ob uporabi predlaganega algoritma. Glede na relativno enostavnost in robustnost algoritma ga lahko vgradimo tudi v sistemih vodenja z nezahtevno procesorsko močjo. Reducing energy demand and consumption in residential buildings is a global challenge. Heating of buildings consums nearly 40 % of global energy production. The largest energy consumers in buildings are the Heating, Ventilation and Air Conditioning (HVAC) systems. For reducing energy consumption, reliable control of HVAC systems is required. When designing HVAC control systems, one should take into consideration a possibile sensor failure. Fault of any component in the control loop can increase the energy consumption, shorten lifespan of the rest of the equipment and usually degrade the living comfort. Single fault of the measurement and control component in the temperature control system leads to unwanted changes in the overall control loop, since controller does not receive accurate sensor data. A general practice during the sensor failure is to ''freeze'' the appropriate controller output. Usually, this is not the most appropriate solution, since it can cause larger temperature deviations in a particular room where sensor failure occurs. The scope of the problem presented is to develop and evaluate fault-tolerant algorithm. Developed algorithm should be simple and suitable for implementaion in practice. In this thesis, an innovative concept is proposed for maintaining high control performance when sensor failure occurs. It is based on the findings that valves openings and indoor temperatures are correlated. Therefore, we introduce the new functions which reflect interacting ratios of the valves openings in adjacent rooms. When sensor failure occurs, the proposed system claculates the control signal from the valve openings of the adjacent rooms. We evaluated the algorithm on simulation-based experiment of building model, developed in MATLAB/Simulink programming environment. In the simulations we used real-world weather data for the city of Ljubljana, for a period of one year, with a sampling rate of one hour. It was shown high correlation between the fault-free system and the system with a single sensor fault while using the proposed algorithm. Since the algorithm is relatively simple and robust, it could be implemented in practice in a control systems with low processing power.

This paper is concerned with the stabilization problem for a class of Markovian stochastic jump systems against sensor fault, actuator fault and input disturbances simultaneously. In the proposed approach, the original plant is first augmented into a new descriptor system, where the state vector, disturbance vector and fault vector are assembled into the state vector of the new system. Then, a novel augmented sliding mode observer is presented for the augmented system and is utilized to eliminate the effects of sensor faults and disturbances. An observer-based mode-dependent control scheme is developed to stabilize the resulting overall closed-loop jump system. A practical example is given to illustrate the effectiveness of the proposed design methodology Refereed/Peer-reviewed

Sensor failures are manifested by a series of events ranging from modest signal drift to severe malfunctioning leading to the eventual replacement of the sensor. This procedure straightforwardly restores the original conditions of the array, but the non-reproducibility among the same kind of sensors as well as their deterioration (drift) during the lifetime may require frequent re-calibrations. In this paper we elaborate the concept that sensor arrays may be rather unsensitive to single elements’ failures, and for the scope a data processing architecture based on a set of sub-classifiers one for each sensor is introduced. The global class assignment is obtained through a weighted majority voting decision rule and the model has been here validated with both artificial and real gas sensor array datasets related to a study case where a fixed number and kind of classes are randomly presented to the sensors. In both cases, some of the sensors were suffering from drift whose magnitude is large enough to deceive multicomponent classifiers, such as KNN. Artificial sensor fault events were also induced in the array to evaluate the fault-tolerance property of the model. Results indicate that both for artificial and real datasets the classification rate achieved by the model largely outperformed that of KNN.

This study focuses on the benefits that can be induced by a Distributed Induction Machine Railway Traction. This is obtained through the additional degrees of freedom it provides, on the one hand across the cooperative controls of the multi-converter/multi-machine systems which constitute the distributed traction and on the other hand by exploiting the Natural Structural Redundancy available in this system. The main objective we want to reach through these two aspects, cooperative controls and structural redundancy, is to ensure service continuity of the distributed railway traction. This study leads to design a supervising automaton which allows ensuring service continuity of the studied system in the case of external disturbances by making the necessary adjustments and drive control commutation. The analysis of cooperative controls leads to the improvement of the Average Differential Control and to propose a new anti-slip strategy. The structure redundancy advantage of a multi-converter/multi-machine system is used in order to compensate the speed sensor fault. Finally, the different strategies which have been proposed are experimentally validated on a test bench that emulates a traction bogie composed by a dual-inverter/dual-motor system.; Cette étude s'intéresse principalement aux avantages que peut offrir une architecture d'une traction ferroviaire électrique asynchrone répartie. Ceci est obtenu à travers les degrés de liberté additionnels qu'elle apporte, d'une part, par la commande des systèmes multi-convertisseurs/multi-machines la constituant, d'autre part, par l'exploitation de la redondance structurelle qu'elle offre naturellement. L'objectif principal qu'on cherche à atteindre au travers de ces deux aspects, commande et redondance structurelle, permet d'assurer la continuité de service de la chaîne de traction répartie en présence de différentes perturbations. Cette étude nous amènera à la conception d'un organe décisionnel qui apporte les adaptations, les reconfigurations de commandes voire d'observateurs nécessaires pour maintenir le bon fonctionnement du système en modes nominal et perturbé. L'analyse des différentes structures de commandes coopératives d'un système bionduleur/bimoteur nous a conduit à l'évolution de la Commande Moyenne Différentielle pour proposer une stratégie d'anti-patinage. Les différentes stratégies ainsi proposées ont été validées expérimentalement sur une plate-forme émulant la charge mécanique d'un bogie de traction composé d'un système bionduleur/bimoteur.

International audience; In this paper, a new method for determining the Principal Component Analysis (PCA) model structure for system diagnosis is proposed. This method, based on the variables reconstruction principle, determines the PCA model optimizing detection and isolation of single or multiple faults affecting redundant or non redundant variables of a system. This new method has been validated by a simulation example.

International audience; A method for failure detection and isolation (FDI) for redundant aircraft sensors is presented. The outputs of the concerned sensors are involved in the computation of §ight control laws, and the objective is to eliminate any fault before propagation in the control loop when selecting a unique §ight parameter among a set (generally, three) of redundant measurements. The particular case of an oscillatory failure is investigated. The proposed method allows an accurate FDI of erroneous sensor and computes a consolidated parameter based on the fusion of data from remaining valid sensors. The bene¦ts of the presented method are to enhance the data fusion process with FDI techniques which improve the performance of the fusion when only few sources (less than three) are initially valid.

International audience; In this contribution we investigate the problem of simultaneous observer based sensor fault reconstruction and state estimation of a class of linear time-varying (LTV) systems that are skew-symmetric models. The main features concern the use of a bank of observers to detect and isolate faulty sensors and in the same time provide unbiased state estimation. On the other hand, we introduced a switching gain technique to deal with singular points. Stability analysis is achieved thanks to the Barbalat's lemma and without solving the well-known time-varying Sylvester equation. The proposed approach is extended to more general LTV systems of any order.

International audience; This article proposes an algebraic method to fault diagnosis for uncertain linear systems. The main advantage of this new approach is to realize fault diagnosis only from knowledge of input and output measurements without identifying explicitly model parameters. Using tools and results of algebraic identification and pseudospectra analysis, the issues of robustness of the pro- posed approach compared to the model order and noise measurement are examined. Numerical examples are provided and discussed to illustrate the efficiency of the proposed fault diagnosis method.

Reliability of sensor measurement is vital to assure the performance of complex and nonlinear industrial operation. In this paper, the problem of designing and development of a data-driven multiple sensor fault detection and isolation (MSFDI) algorithm for nonlinear processes is investigated. The proposed scheme is based on an evolving multi-Takagi Sugeno framework in which each sensor output is estimated using a model derived from the available input-output measurement. Our proposed MSFDI algorithm is applied to continuously stirred tank reactor sensor fault detection and isolation. Simulation results demonstrate and validate the performance capabilities of our proposed MSFDI algorithm. 2012 IEEE. Qatar National Research Fund Scopus

International audience; This paper deals with the problem of sensor fault tolerant control for Takagi-Sugeno nonlinear systems. Firstly, a residual generator is designed in order to detect and isolate sensor faults. Secondly, a nonlinear observer based controller, adopting the so-called parallel distributed compensation (PDC) structure is designed. The PDC controller is based on a weighted blending of the estimated states provided by different observers. Each observer is constructed to estimate the system state from the inputs and only one output. The blending of the estimates depends on the residual magnitudes in order to minimize the use of faulty estimates in the control law. The stability of the global closed-loop system is studied by Lyapunov theory and the gains of the fault tolerant controller are obtained by solving Linear Matrix Inequalities (LMI). Numerical examples are provided.

International audience; This paper deals with the problem of fault tolerant control of nonlinear systems represented by Takagi-Sugeno models subject to sensor faults. Observer based controllers are designed for each faulty-situation (mode). The classical switching law is replaced by a new mechanism which avoid the switching phenomenon. The purpose is to be able to study the stability of the global closed-loop system. This new mechanism uses the residual signals obtained by a residual generator. A bank of observers is designed and each observer uses only one output. Each observer based controller is designed using the estimated state provided by the corresponding observer. Finally, the control law is constructed from these different controllers by using smooth weighting functions depending on the residual signals and satisfying the convex sum property. This last allows to study the stability of the closed-loop system by Lyapunov theory and the tools developed for Takagi-Sugeno systems. Linear Matrix Inequality (LMI) conditions are then proposed to ease the design of a such fault tolerant controller.

This work contributes to develop a new algebraic and deterministic approach of fault diagnosis, which is not based on an a priori explicit model of the process and presents a new perspective based on the theory of distributions and the pseudospectra analysis of matrix pencils. Our approach is based on certain tools and developments in the theory of algebraic estimation usual in automatic signal processing communities. The considered systems form a particular class of uncertain linear dynamical systems. The treated faults are actuator and sensor faults modeled through structured signals. The main advantage of this new approach is to realize fault diagnosis only from knowledge of input and output measurements and without identifying explicitly model parameters. The method exploits the fact that a structured signal satisfies a differential equation. The decision is entirely based on the temporal evolution of the estimates of some constant features of the fault. The effectiveness of the proposed approach is illustrated through academic examples., Ce travail contribue à développer une nouvelle approche algébrique et déterministe de diagnostic qui ne nécessite pas la connaissance a priori d'un modèle explicite du système. Il présente un nouveau point de vue basé sur la théorie des distributions et l'analyse du pseudospectre de faisceaux de matrices. Notre démarche s'inspire de certains outils et développements de la théorie de l'estimation algébrique courants en automatique et traitement du signal. Le principal atout de cette approche est qu'il est possible, sous certaines hypothèses, de détecter, localiser et identifier les défauts à l'aide des seules mesures de la commande et de la sortie sans avoir à estimer les paramètres du modèle. Nous considérons que les défauts actionneur et capteur sont modélisés par des signaux structurés. La méthode exploite le fait qu'un signal structuré satisfait une équation différentielle. La prise de décision est entièrement fondée sur l'évolution temporelle des estimations de certaines caractéristiques constantes des défauts. L'efficacité de l'approche présentée est illustrée au travers d'exemples académiques.

This paper investigates the problem of sensor fault estimation and fault-tolerant control for Markovian jump systems with time delay and Lipschitz nonlinearities. The issues involved here are: i) sensor faults; ii) model Lipchitz nonlinearities; iii) system structure changes governed by Markovian jumping parameters; and iv) time delay in system states. Such type of mathematical models can represent a large number of practical systems in the actual engineering. A new estimation technique (named proportional and derivative sliding mode observer) is developed to deal with this design problem. The proposed observer is mode-dependent type in which a derivative gain and a proportional gain are introduced to provide more design freedom, and a discontinuous input term is introduced to eliminate the effects of sensor faults. By employing the developed estimation technique, the asymptotic estimations of system states and sensor faults can be obtained simultaneously. Based on the estimation, an observer-based fault-tolerant control scheme is developed to stabilize the resulting closed-loop system. Finally, a numerical example is presented to illustrate the effectiveness and applicability of the proposed technique. Refereed/Peer-reviewed

International audience; This article proposes an algebraic method to fault diagnosis for uncertain linear systems. The main advantage of this new approach is to realize fault diagnosis only from knowledge of input and output measurements without identi- fying explicitly model parameters. Using tools and results of algebraic identification and pseudospectra analysis, the issues of robustness of the proposed approach compared to the model order and noise measurement are examined. Numerical examples are provided and discussed to illustrate the efficiency of the proposed fault diagnosis method.

International audience; This paper deals with sensor fault detection within a reconfigurable direct torque control of an induction motor-based electric vehicle. The proposed strategy concerns current, voltage and speed sensors faults that are detected and followed by control reconfiguration in order to allow the vehicle continuous operation. The proposed approach is validated through experiments on an induction motor drive and simulations on an electric vehicle using a European urban and extra urban driving cycle.

International audience; We propose a diagnosis approach of sensor and actuator modeled as structured signals acting on a particular class of uncertain linear dynamical systems. The main advantage of this approach is that it is possible under certain assumptions, to detect, isolate and identify faults using only input and output measurements without having to identify model parameters. The method is based on the generation and analysis of analytical redundancy relations and exploits the fact that a structured signal satisfies a differential equation. The decision rule is based entirely on the temporal behaviour of the estimates of some fault characteristics. A numerical example is provided and discussed to illustrate the proposed approach.; Nous proposons une méthode de diagnostic des défauts actionneur et capteur modélisés au travers de signaux structurés sur une classe particulière de systèmes dynamiques linéaires incertains. Le principal atout de cette approche est qu'il est possible, sous certaines hypothèses, de détecter, localiser et identifier les défauts à l'aide des seules mesures de la commande et de la sortie sans avoir à identifier les paramètres du modèle. La méthode est fondée sur la génération et l'analyse de relations de redondance analytique et exploite le fait qu'un signal structuré satisfait une équation différentielle. La prise de décision est entièrement fondée sur l'évolution temporelle des estimations de certaines caractéristiques des défauts. Un exemple numérique est fourni et commenté afin d'illustrer l'approche proposée.

International audience; This technical paper deals with the transition performance improvement of a sensor fault-tolerant controller devoted to Electric (EV) or Hybrid Electric Vehicles (HEV). Indeed, improvements are brought over a previously developed technique that exhibit abrupt changes in the torque if a sensor fault is detected and after a transition from a control technique to another one [1]. The Fault-Tolerant Control (FTC) system firstly concerns the sliding mode control technique since better performances are obtained with an encoder to get the speed information. In the event of unavailability of the speed sensor, a sensorless fuzzy control technique is applied. In the proposed active fault-tolerant control approach a short and a smooth transition are achieved from the encoder-based control technique to the sensorless one using an appropriate fuzzy logic decision approach.

International audience; This paper deals with the problem of fault detection and identification in noisy systems. A proportional integral observer with unknown inputs is used in order to reconstruct state and sensors faults. A mathematical transformation is made to conceive an augmented system, in which the initial sensor fault appear as an unknown input. This reconstruction is made by the use of a proportionnel integral observer. The noise effect on the state and fault estimation errors is also minimized. The obtained results are then extended to nonlinear systems described by nonlinear Takagi-Sugeno models.

This paper presents Confidence, a tool for identifying and addressing faults in wireless sensing systems. Confidence pinpoints potential sensor and network faults in real time, allowing users to validate unexpected data and address any failures in the field. By introducing a well defined, low-dimension feature space, and functions to map sensor data into this space, we are able to achieve fault detection and diagnosis with relatively simple mechanisms such as outlier detection. Users can directly modify system outcomes by altering a classification label in instances when Confidence's automated algorithm draws the wrong inference. This label is applied to all similar points in the feature space, enabling Confidence to learn from user interaction in the field. This abstraction for incorporating user knowledge provides a lightweight and easy-to-understand interface for the user, while limiting user burden and reducing the required a priori environmental knowledge. Confidence has performed well on real-world deployments, including one deployment of 130 sensors, replayed datasets, and network simulations. Confidence accurately detects and diagnoses at least 90% of all data, and user interaction improves it's performance.

Diesel propulsion plant for large ships is very complex, high risk process with strong demands on its reliability and availability. The most important information and data about plant dynamic behaviours during normal and faulty operation needed for diagnostic and control system are collecting from sensors, so their fail-operational feature is of very importance to overall plant reliability. An effective scheme in sensor faults detection and identification within marine diesel propulsion plant using on-line estimation method is given in the paper. The noise in sensor signals is taken into consideration. Simulation example of diesel engine speed sensor fault detection method based on measured and estimated signal values is presented. The Kalman filtering is used. Simulation results have shown good effectiveness of proposed scheme in engine speed sensor faults detection.

In this paper, we present a data processing architecture based on a set of cooperative classifiers aimed at class recognition with sensor arrays. The main property of the method is the mitigation of the performance degradation due to drift and fault of individual sensors. Moreover the capability to preserve high classification rate even with a faulty sensor allows replacing and calibrating a sensor during the measurement itself. The algorithm has been positively tested with an experimental dataset with an array of commercial metal oxide semiconductor sensors used to identify a number of volatile compounds.